CIMMYT’s Strategy for Catalyzing the Adoption of Conservation Agriculture in
Southern Africa.
Brief history of Conservation Agriculture
Conservation Agriculture (CA) is a term that has been coined in the last 10 years to
describe agricultural systems that include three major components: minimal soil
movement, permanent cover of the soil surface generally with crops and crop residues,
and crop rotation. The term has
benefits in that it implies a system
approach and takes the emphasis off
the soil tillage component implicit in
terms such as conservation tillage
(which describes different although
overlapping systems) and direct
seeding, no-tillage or zero tillage
(previous names for CA which
focused on the tillage or lack of it).
In CA systems major benefits accrue
Conservation Agriculture plot in Hwedza,
from all three major components and
Zimbabwe. Soybeans seeded after maize
the interactions between them, and in
(crop rotation) without soil tillage and with
many cases it has been shown that
good
residue cover from maize residues and
omitting one or more of the
thatch grass.
principles results in unsustainable
systems.
Most of the early development of CA concentrated on large, mechanized commercial
farms and was driven initially by the interests of agrochemical companies (ICI and later
Monsanto), but then later by farmers and farmer associations as they realized the major
economic and conservation benefits of CA.
In the 1980’s and 1990’s efforts began to extend the practice of CA to resource-poor
smallholder farmers, notably in Brazil, Paraguay, Ghana, India and Pakistan. In the first
three of these countries there is now extensive use of CA by small-scale farmers, while in
India and Pakistan there are large areas of wheat in the rice-wheat system sown using CA
principles, but then the rice crop is seeded using conventional agriculture.
In Southern and Eastern Africa there have been numerous efforts to adapt CA principles
to smallholder farmer situations. The German GTZ has had projects in several countries
in the region helping develop and extend CA systems, notably in Tanzania, Mozambique
and Zimbabwe. FAO has more recently been involved in Kenya, Tanzania, Mozambique
and Zimbabwe. Sasakawa Global 2000 initiated some efforts in Malawi. However, the
biggest developments in CA have probably been in Zambia and more recently in
Zimbabwe. The Zambian efforts originated with the experiences of a commercial
Zimbabwean farmer, Brian Oldrieve. Based on his experiences with small-holder CA
production systems in Zimbabwe, the Zambian National Farmers Union created a
Conservation Farming Unit which developed and promoted the local system, largely
based on preparation during the dry season of planting basins into which the crops are
sown each season. Basins only cover 5-15% of the land area and so tillage is markedly
reduced. Haggeblade and Tembo (2002) estimated that there were about 40,000 farmers
in Zambia practicing the system on at least part of their land. Together with the Golden
Valley Agricultural Research Trust (GART), the CFU have developed and extended an
animal traction CA system, although adoption of this system has been slight. Based on
the success of the basin system in Zambia, the British DFID catalyzed the re-introduction
of the system to Zimbabwe in 2004, funding a number of NGOs under an FAOcoordinated Conservation Farming Task Force. Under this Protracted Relief Project a
very large number of farmer-managed plots have been installed over the last three
seasons in Zimbabwe, concentrating on the “poorest of the poor” managing manual
systems. Both the Zambia CFU and the DFID/FAO work in Zimbabwe have been
development efforts, with some research to support it – especially the GART work in
Zambia. The approach has tended to be prescriptive rather than adaptive, although
farmers have been encouraged to adapt the system after they have learned how to manage
it. Inputs were supplied to farmers for the CA plots in the first years.
It was in this context that CIMMYTs CA project for southern Africa began. The project
had been in preparation for several years, but was funded and initiated in 2004. There
were several conceptual pillars upon which the project was built apart from the three
basic principles of CA. These concepts, a result of CIMMYTs close involvement with
CA work in South Asia and Latin America, included the following:
• Mind-set and the culture of the plough are two of the major impediments to the
expansion of CA
• CA is the most sustainable system we know of at the moment for field scale
agriculture.
• While the three principles of CA are extremely widely applicable, the techniques and
technologies to apply these principles are dependent on local biophysical
conditions and farmer circumstances.
• Research and extension systems do not have the capacity to develop complex systems
for all farmers
• Crop residue retention is vital for the success of CA. Although residue retention is
difficult when farmers use all residues for animal feed, the alternative of declining
soil organic mater levels is not sustainable.
Facilitating the Widespread Adoption of Conservation Agriculture in Southern
Africa – CIMMYTs Projects.
History:
• July 2004. BMZ-funded project begins in Malawi, Tanzania, Zambia and Zimbabwe,
with field activities initiated in all four countries in the 2004/2005 crop season.
• April 2006. CIAT-led project, in which CIMMYT plays a major role for field
activities in CA, is approved by the Sub-Saharan Africa Challenge Program to
work in the Zimbabwe-Malawi-Mozambique Pilot Learning Site.
• December 2006. Small grant received from German GTZ for a workshop on current
status of CA in Mozambique. The workshop was held the same month.
•
March 2007. CIAT-led project, in which CIMMYT plays a major role for field
activities in CA, is approved by the Austrian Government for work in Manica and
Sofala Provinces of Mozambique.
• June 2007. IFAD agrees to a grant to fund CIMMYT CA project in Malawi,
Mozambique, Zambia and Zimbabwe, effectively carrying on from the BMZ
project.
The following paragraphs describe the different interlinked activities carried out under
the CA Projects in which CIMMYT is involved in Southern Africa and the rationale
behind them.
Definition of pilot sites and “best bet” CA technologies with partners.
An initial workshop was held in the region with the main partners identified before the
project began, including partners from Malawi, Mozambique, Tanzania, Zambia,
Zimbabwe, FAO, DFID, GTZ and ACT (African Conservation Tillage Network). Based
on this workshop a greater number of interested stakeholders/potential partners were
identified in each of the project countries. A Stakeholder workshop was then held in each
country to define the pilot sites where the project would work and the “best bet” CA
technologies based on previous work on CA and associated technologies in each region
of each country.
Conducting the baseline survey in
Karatu, Tanzania
Baseline surveys
Baseline surveys were conducted in most of the
initial communities where the project began field
activities in the 2004/2005 crop season. Baseline
surveys explored current farmer practices and
problems, wealth endowment, and exposure to
conservation agriculture techniques. Since then field
activities have expanded into “new” communities,
but baseline surveys have not been conducted in all
of these communities.
Community Awareness and Farmer Discussion
Groups.
Before field activities are initiated in each
community, an awareness meeting is held with
members of the community to a) find out from them
what their agricultural production problems are; b)
explain to them the principles of CA and link the
benefits of CA to the problems that they have
enunciated (normally there is a large degree of
convergence); c) explain that we would like to help
them manage some demonstration plots in the
community; and d) ask them to decide on whose
fields in the community the demonstration plots
should be established.
Farmers observe and discuss
signs of soil degradation in
Shamva, Zimbabwe
Demonstration and Validation Plots.
Simple demonstration sites have been
established in all of the pilot
communities. These plots are established
each year on the same site to assess the
cumulative effects of CA over time. We
aimed at 5-6 sites in a community, each
comprising 3 plots: normal farmer
practices and management and two
different CA options (best bets). All
plots receive the same level of fertilizer
and are seeded to the same variety to
A demonstration plot in Balaka, Malawi
take these two variables out of the
in the 2004/05 season with the farmer’s
equation. The actual CA options vary
check on the left and the “best bet” CA
from country to country, from areas with
practice on the right.
manual systems to areas with animal
traction systems, and also depending on soil type and rainfall pattern. We have had a
problem convincing partners that the demonstration plots should be confined to one
community and that they should be close to each other, a practice we feel is important as
it allows participating farmers to visit all the plots easily and also for them to feel “group
support”. However, partners have often spaced the plots out more than we would have
liked, given normal extension practices of widespread blanket recommendations.
While for the public and NGO partners with whom we work the plots are for
demonstration to farmers, to the project they are “validation” trials of the best bet
technologies (which may be modified somewhat from year to year based on observations
and farmer perceptions) with 5-6 replications in the community. This allows us to analyze
the results statistically and accrue valuable, scientifically relevant information on the
performance of CA practices on farmers’ fields.
Farmer group fertilizing their on-farm
demonstration plot in Enyezini, Malawi.
Farmers are encouraged to work in groups
in order to increase the exposure of
farmers to the plots. Farmers receive all
the inputs at the start of the season, but at
the end of the season return enough grain
to cover the costs of the inputs to a
community project. In this way farmers do
not participate just to get free inputs, the
economic benefits of the technology are
more evident, and (as they do not pay us
for the inputs) we do not create the
impression that the project is trying to
make money out of their efforts.
A summary of the number of pilot communities in each of the countries over the past four
seasons is shown in the following table.
Table 1. Numbers of pilot communities with CA demonstration activities in five countries over the
period 2004-2008.
Country
2004/05 2005/06 2006/07 2007/08
Malawi
3
5
8
10 a
Mozambique b
4
26 c
d
Tanzania
2
3
2
Zambia
2
2
2
4
Zimbabwe
2
4
5
6
Total
9
14
21
46
a
Includes two communities managed under the SSA Challenge Program
project
b
Mozambique was not included in the BMZ/GTZ project
c
Includes 2 communities under the SSA-CP project and 20 communities
under the CIAT-Austria project.
d
Tanzania is not included in the IFAD-funded project.
Capacity building of partners
At the start of the project, many of the partners
had little or no exposure to the principles and
practice of CA. Conviction of the necessity and
benefits of CA only comes with practical field
experience, but it is important for partners to
have a basic grounding in the principles and
possibilities of CA. Therefore we conducted a
series of short, 3-day courses for partners in
each of the project countries. To date six
courses have been conducted (one each in
Malawi, Mozambique, Tanzania and Zambia,
and two in Zimbabwe).
An on-farm research trial on
weeding intensity and
fertilizer/manure amounts in
Monze, Zambia
Partners learn to use, calibrate and
maintain an animal traction seeder
during the short CA training
course in Zambia.
Research on knowledge gaps.
Knowledge gaps for successful CA management were
identified at the start of the project in some of the
project countries, whereas in others (e.g. Zimbabwe)
there was no consensus among stakeholders on CA
knowledge gaps/researchable topics. Therefore, onfarm research to identify solutions to specific CA
management problems was initiated in some countries
in the first season, while in others these activities
started later. Issues for research have since been
identified with partners in study tours of on-farm demonstration sites. These are
discussed at Annual Evaluation and Planning Meetings (see below) and partners
identified to manage the on-farm research activities. Occasionally it has been decided that
on-station research is preferable or acceptable, and trials have then been conducted on
public research stations.
Equipment testing, modification and manufacture
Adequate equipment has been a severe
impediment to successful research,
extension and adoption of CA in all types
of traction systems: manual, animal and
mechanical. For manual systems very
basic equipment (e.g. pointed stick) can be
used to successfully manage CA, whereas
in animal or mechanical systems the
equipment is more complicated and
expensive. There has been relatively little
development of adequate equipment for
managing CA in manual or animal traction
A farmer tries out a Zimplow seeder
systems, but in Brazil there are now
at a participatory (farmers and
several successful companies
machinery manufacturers) field day
manufacturing such equipment. Initially,
organized by the CA project.
therefore we imported both manual “jab
planters” and animal traction direct seeders form Brazil to test in the region. We have
also used a locally produced intermediate technology for animal traction systems – a
ripper tine that is used to open a furrow into which seed and fertilizer are placed manually.
As continued importation of equipment will be difficult and expensive we have
endeavored to catalyze and promote local manufacture of adapted equipment, and have
had jab planters made by two companies in Zimbabwe, as well as testing similar planters
manufactured in Tanzania. One company in Zimbabwe has also manufactured an animal
traction direct seeder. Field days have been held with participation of farmers and
machinery manufacturers to test and discuss
equipment. However, the local equipment
manufactured to date suffers from marked
quality problems, and while we will
continue to support local manufacture we
may continue to promote importation of
robust equipment.
Farmer discussion groups.
Partners are encouraged to hold regular
discussion with interested farmers (including
the farmers and farmer groups in the
community who manage the demonstration
plots) to observe and discuss the
development of the crops in the
A farmer discussion group hosted by
Mr. Makwara in Zimuto, Zimbabwe.
demonstration plots and in any “farmer experiments” 1 and learn from these experiences.
In many ways these discussion groups are similar to the concept of the Farmer Field
School, especially where there may be some technical input on specific aspects.
After each season partners conduct a farmer discussion group meeting in each community
to report back on the results of the demonstration plots in the community, and to discuss
these with the farmers. Perceptions of farmers on CA and CA management are also
discussed, and these points provide inputs into discussions at the Annual Evaluation and
Planning Meetings.
Annual evaluation and planning meetings
These meetings are held in each country each year between crop seasons and attended by
all the technical partners involved in the project. Farmers do not attend these meetings,
but their suggestions and perceptions collected during the discussion groups in each
community provide inputs for the discussions and decisions. Typically an annual meeting
will comprise: presentations and discussions of field results from each community and
from research trials; enunciation and analysis of the technological and logistical problems
that occurred during the previous season; analysis and definition of solutions to the
problems defined; definitions of roles and responsibilities, taking into account the
agreements on functional solutions to problems; definition of budgets for the next season
for each community and each research partner and/or activity; definition of
demonstration plot management, sites and communities for the coming season; definition
of on-farm and on-station research trials; and, finally, preparation of trial and
demonstration plot protocols that incorporate all the decisions made. Partners are
furnished with these protocols and data sheets for each site.
Participants in the 2008 study tour in
Zambia pose in a demonstration plot in
Kabwe.
1
Partner Study Tours
Each season we tour field sites with all of
the technical project partners in each country,
observing trials and demonstration plots,
meeting and discussing CA issues with
farmers, and identifying issues that need to
be resolved or that offer opportunities for
research in the future. These tours serve also
to stimulate project partners, and provide
valuable inputs into the evaluation and
planning process. Initially we were able to
visit all sites in each country, but as the
project has expanded, for budgetary reasons
we are now unable to do so. However, each
year we visit at least one geographical area
in each country.
We term any field or part of a field where a farmer is trying/testing CA practices under his/her own
initiative a farmer experiment. We believe that these farmer experiments precede technology adoption.
Monitoring of farmer experimentation
On the development side of the project,
the main aim of the demonstration
plots is to stimulate farmer interest,
experimentation and adoption of CA
technologies. At the same time it is
common for farmers to modify and
adapt technologies to better fit their
particular circumstances. This gives an
opportunity to researchers and
extension agents to observe and learn
from these farmer modifications, and
possibly incorporate them into the
research and extension programs.
Therefore, together with partners, we
try to identify, georeference and
monitor these farmer experiments.
Mrs. and Mr. Tafireyi of Zimuto,
Zimbabwe, stand proudly in their
conservation agriculture experiment plot.
Farmer study visits
As noted earlier, the project works in a few pilot sites in each country. Scaling out of the
technologies to other sites and communities is being stimulated through farmer study
visits. Starting in 2007, groups of farmers from four communities (two in Malawi and
two in Zimbabwe) were taken on a 2-day visit to the project communities. The principal
interaction was between the farmers themselves, with technical staff present for part of
the time to help answer technical issues. However, the extension agent with responsibility
for the community of the visiting farmers also visited the extension agents of the pilot
community to discuss and learn about CA.
In Malawi, at the insistence of local partners, we have agreed to help establish
demonstration plots in both of the “visiting” communities, although we plan to have less
involvement there in the future so that we can monitor the effectiveness of the study visit
in stimulating farmer experimentation with CA. In Zimbabwe however, we have initiated
demonstration plots in only one of the two communities, which will allow us to monitor
differences in farmer experimentation. Although the sample is very small, we hope to get
some insights on the effectiveness of the study visits and demonstration plots that will
help design future activities.
Case studies/Innovation histories
It is unlikely that in the life of relatively short-term 3-year projects there will be
considerable adoption of CA, even in the target communities. It is clear from experiences
in other parts of the world that the initial stages of adoption and consolidation of CA
systems in a community or region takes many years, after which growth can be
exponential. Therefore, we do not plan farmer surveys within the life of the present
project, but rather case studies of experimenting farmers to obtain information on their
and their family’s perceptions of CA, any impacts they have observed to date on their
food security, income and/or livelihoods, and their views on the potential impacts of CA
on these. Some initial case studies and innovation histories of the farm families have been
conducted in 2007.
Researcher-managed medium-term trials
It is important to be able to observe and understand the longer-term effects of CA on crop
productivity and soil quality. This information provides inputs for scaling out activities in
that it helps predict the impacts of widespread CA adoption. These predictions can be
made simply on an area basis taking into account
different recommendation domains and/or geographically
and climatically similar areas, or through incorporation
of the data into crop/soil simulation models also linked to
geographic information systems. While these data could
in theory be obtained from the demonstration plots in
farmers fields in the region there are many issues of lessthan-optimum management, farmer decisions on land
allocation changing over time (impacting on the
continuity of the demonstration plots), and damage or
loss of scientific equipment due to marauding animals or
theft. Therefore, we established two “long-term” trials 2
at the start of the BMZ project in 2004: one in Choma,
Zambia, and one at Henderson Research Station,
Taking soil moisture data on Zimbabwe. After the first season we suspended the
the conventionally tilled plot Zambian trial due to some problems of soil variability
of the long-term trial at
and trial management, and started a new trial in Monze,
Henderson Research
Zambia. Under the CIAT-Austria project we initiated a
Station, Zimbabwe. The
similar trial in Chimoio, Mozambique, in the 2006/2007
run-off and erosion plot can
season, and at Chitedze Research Station, Malawi, under
be seen in the background.
the IFAD project in the present crop season.
Ph.D., M.Sc. and undergraduate thesis studies.
Under the previous BMZ-funded project, two Ph.D. scholarships were awarded to
students from the southern Africa region to study at German universities, with their field
work conducted in southern Africa in association with the CA project. The disciplines for
these scholarships were selected given their expected importance in determining the
expansion of CA and the availability (or lack of availability) of expertise in the region,
and therefore one scholarship was given for socio-economic studies and one on soil
ecology. The soil ecology Ph.D. student successfully defended his thesis in December
2007, and the socio-economic student defended his thesis in late January 2008.
Although full-time scholarships are not included in the IFAD project, we help and
encourage undergraduate and post-graduate students to conduct their thesis work on CA
project activities. At the moment in Zimbabwe we are working with one M.Sc. student
2
Long-term reflects optimism that these trials will continue until the cumulative effects of CA on soil
quality and crop productivity are clear and documented. However, short-term project funding and the
limited budgets of national programs in the region mean that we can only guarantee that these trials
continue until the end of the presently funded projects.
and one undergraduate student from the University of Zimbabwe, while in Mozambique,
one Ph.D student and one M.Sc. student are conducting their fieldwork on the long-term
trial and a Ph.D. student is working on the economics of smallholder CA, all under the
CIAT-Austria project.